In this thesis, we consider multiobjective combinatorial optimization problems. We address two main topics. We first address the polynomially solvable cases of the Traveling Salesperson Problem and the Bottleneck Traveling Salesperson Problem. We consider multiobjective versions of these problems with different combinations of objective functions, analyze their computational complexities and develop exact algorithms where possible. We next consider generating extreme supported nondominated points of multiobjective integer programming problems for any number of objective functions. We develop two algorithms for this purpose. The first one is an exact algorithm and finds all such points. The second algorithm finds only a subset of extreme supported nondominated points providing a worst case approximation for the remaining points.

Multivariate adaptive regression spline (MARS) denotes a modern methodology from statistical learning which is very important in both classification and regression, with an increasing number of applications in many areas of science, economy and technology. MARS is very useful for high dimensional problems and shows a great promise for fitting nonlinear multivariate functions. MARS technique does not impose any particular class of relationship between the predictor variables and outcome variable of interest. In other words, a special advantage of MARS lies in its ability to estimate the contribution of the basis functions so that both the additive and interaction effects of the predictors are allowed to determine the response variable. The function fitted by MARS is continuous, whereas the one fitted by classical classification methods (CART) is not. Herewith, MARS becomes an alternative to CART. The MARS algorithm for estimating the model function consists of two complementary algorithms: the forward and backward stepwise algorithms. In the first step, the model is built by adding basis functions until a maximum level of complexity is reached. On the other hand, the backward stepwise algorithm is began by removing the least significant basis functions from the model. In this study, we propose not to use the backward stepwise algorithm. Instead, we construct a penalized residual sum of squares (PRSS) for MARS as a Tikhonov regularization problem, which is also known as ridge regression. We treat this problem using continuous optimization techniques which we consider to become an important complementary technology and alternative to the concept of the backward stepwise algorithm. In particular, we apply the elegant framework of conic quadratic programming which is an area of convex optimization that is very well-structured, herewith, resembling linear programming and, hence, permitting the use of interior point methods. The boundaries of this optimization problem are determined by the multiobjective optimization approach which provides us many alternative solutions. Based on these theoretical and algorithmical studies, this MSc thesis work also contains applications on the data investigated in a T&Uuml / BiTAK project on quality control. By these applications, MARS and our new method are compared.

In Multi-Skill Project Scheduling Problem (MSPSP) with hierarchical levels of skills, there are more than one skill type and for each skill type there are levels corresponding to proficiencies in that skill. The purpose of the problem is to minimize or maximize an objective by assigning resources with different kinds of skills and skill levels to the project activities according to the activity requirements while satisfying the other problem dependent constraints. Although single-objective case of the problem has been studied by a few researchers, biobjective case has not been studied yet. In this study, two objectives, which are the makespan and the total skill wasted, are taken into account and while trying to minimize the makespan, minimizing the total skills wasted is aimed. By the second objective, overqualification for the jobs is tried to be minimized in order to prevent job dissatisfaction. The biobjective problem is solved using a Multiobjective Genetic Algorithm, NSGA-II. The results of the proposed algorithm are compared with the GAMS results for small-sized problems and with the random search for larger problem sizes.

The thesis presents research on multiobjective optimization based on memetic computing and its applications in engineering. We have introduced a framework for adaptive multiobjective memetic optimization algorithms (AMMOA) with an information theoretic criterion for guiding the selection, clustering, and local refinements. A robust stopping criterion for AMMOA has also been introduced to solve non-linear and large-scale optimization problems. The framework has been implemented for different benchmark test problems with remarkable results. This thesis also presents two applications of these algorithms. First, an optimal image data hiding technique has been formulated as a multiobjective optimization problem with conflicting objectives. In particular, trade-off factors in designing an optimal image data hiding are investigated to maximize the quality of watermarked images and the robustness of watermark. With the fixed size of a logo watermark, there is a conflict between these two objectives, thus a multiobjective optimization problem is introduced. We propose to use a hybrid between general regression neural networks (GRNN) and the adaptive multiobjective memetic optimization algorithm (AMMOA) to solve this challenging problem. This novel image data hiding approach has been implemented for many different test natural images with remarkable robustness and transparency of the embedded logo watermark. We also introduce a perceptual measure based on the relative Rényi information spectrum to evaluate the quality of watermarked images. The second application is the problem of joint spectrum sensing and power control optimization for a multichannel, multiple-user cognitive radio network. We investigated trade-off factors in designing efficient spectrum sensing techniques to maximize the throughput and minimize the interference. To maximize the throughput of secondary users and minimize the interference to primary users, we propose a joint determination of the sensing and transmission parameters of the secondary users, such as sensing times, decision threshold vectors, and power allocation vectors. There is a conflict between these two objectives, thus a multiobjective optimization problem is used again in the form of AMMOA. This algorithm learns to find optimal spectrum sensing times, decision threshold vectors, and power allocation vectors to maximize the averaged opportunistic throughput and minimize the averaged interference to the cognitive radio network. / February 2016

Made available in DSpace on 2014-06-11T19:30:50Z (GMT). No. of bitstreams: 0 Previous issue date: 2008-12-19Bitstream added on 2014-06-13T20:21:17Z : No. of bitstreams: 1 escobarzuluaga_ah_dr_ilha.pdf: 1508525 bytes, checksum: b6e7b58056f84298f2b063ead5371a59 (MD5) / Fundação de Ensino Pesquisa e Extensão de Ilha Solteira (FEPISA) / O principal objetivo deste estudo é realizar uma análise de aspectos críticos que surgem na modelagem matemática do problema de planejamento da expansão de sistemas de transmissão a longo prazo, assim como o desenvolvimento de ferramentas computacionais para a prova de novos modelos e metodologias que possam contribuir na solução do problema de planejamento de sistemas de transmissão de energia elétrica considerando as condições dos sistemas modernos de energia elétrica. Com esta metodologia, busca-se obter uma rede de transmissão mais eficiente, e com o menor custo possível, que se adapte as novas exigências produzidas pela introdução da desregulação nos sistemas elétricos. Para isto combinam-se três aspectos: rede futura livre de congestionamento, desplanificação e incerteza na geração e na demanda futura, os quais são manipuladas desde a perspectiva mono-objetivo e multiobjetivo. A possibilidade de eliminar completamente o congestionamento na rede de transmissão é analisada através da inclusão no modelo de todos os cenários de geração factíveis futuros, e não somente alguns cenários como outros estudos. Considerar uma operação sem congestionamento para o futuro está associado a grandes custos de investimento. Para atenuar este grande custo uma opção é incluir a possibilidade de desplanificação e a inclusão dos efeitos das incertezas presentes na geração e na demanda futura no problema de planejamento. O problema de planejamento de sistemas de transmissão é um problema de programação não linear inteira mista (PNLIM) quando é usado o modelo DC. Praticamente todos os algoritmos usados para resolver este problema utilizam uma sub-rotina de programação linear (PL) para resolver problemas de PL resultantes do algoritmo de solucão do problema de planejamento, os quais são denominados... / The main goal for this study is to do an analysis of the critical issues that appear in the mathematical modeling of the transmission system expansion planning problem, when long term is considered. A methodology was developed and a computational tool, to solve the transmission expansion planning in modern electrical systems. With this methodology more efficient electrical networks are obtained, at low investment costs. This is accomplished taking into account three important aspects: open access, or congestion-free planning, uncertainty in demand and generation, and de-planning. The problem is solved using mono-objective and multi-objective methodologies. For this investigation, congestion-free transmission networks should consider all the future and feasible scenarios of generation, unlike some papers, where only a few scenarios are taken in to account. This feature is associated to high investment costs. Lower costs are often obtained by the inclusion of uncertainty in future demand and future generation. The transmission system expansion planning problem is a no-linear integer-mixed programming problem (PNLIM) when the DC model is used. Practically, all the algorithms used in the solution process, for this problem, use one subroutine of linear programming (PL) for solved the PL problems that result during the solution process, in the denominated operative problem. The solution of the PL’s is the part of the problem that requires the biggest computational effort, because during the solution process is necessary to solved thousands or millions of PL’s, for high size problems. the PNLIM problem is solved through the combination of a meta-heuristic method and a linear programming method. The meta-heuristic method solves the denominated investment problem and the PL the denominated operational problem. The transmission planning problem considering multiples generation scenarios... (Complete abstract click electronic access below)

Made available in DSpace on 2014-12-17T15:48:07Z (GMT). No. of bitstreams: 1 RomeritoCA_DISSERT.pdf: 1649773 bytes, checksum: 9a9fd0e3782657fe6d014020cdc8fb90 (MD5) Previous issue date: 2013-02-01 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior / This work presents a algorithmic study of Multicast Packing Problem considering a multiobjective approach. The first step realized was an extensive review about the problem. This review serverd as a reference point for the definition of the multiobjective mathematical model. Then, the instances used in the experimentation process were defined, this instances were created based on the main caracteristics from literature. Since both mathematical model and the instances were definined, then several algoritms were created. The algorithms were based on the classical approaches to multiobjective optimization: NSGA2 (3 versions), SPEA2 (3 versions). In addition, the GRASP procedures were adapted to work with multiples objectives, two vesions were created. These algorithms were composed by three recombination operators(C1, C2 e C3), two operator for build solution, a mutation operator and a local search procedure. Finally, a long experimentation process was performed. This process has three stages: the first consisted of adjusting the parameters; the second was perfomed to indentify the best version for each algorithm. After, the best versions for each algorithm were compared in order to identify the best algorithm among all. The algorithms were evaluated based on quality indicators and Hypervolume Multiplicative Epsilon / O presente trabalho apresenta um estudo algor?tmico do Multicast Packing Problem levando em considera??o uma abordagem multiobjetivo. Para tal, faz-se uma extensa revis?o sobre o problema em quest?o. Esta revis?o serviu como ponto de refer?ncia para defini??o de um modelo matem?tico multiobjetivo, tendo em vista que n?o h? na literatura nenhum trabalho que tenha tratado o tema neste aspecto. Em seguida, define-se os casos de teste utilizados no processo de experimenta??o dos algoritmos. Uma vez que tanto modelo matem?tico multiobjetivo quanto os casos de teste foram criados, ent?o desenvolve-se v?rios algoritmos com base nas abordagens cl?ssicas para problemas de otimiza??o multiobjetivo: NSGA2 (3 vers?es) e SPEA2 (3 vers?es). Al?m disso, adaptou-se a metaheur?stica GRASP (2 vers?es) para aplica??o considerando o modelo proposto. Estes algoritmos foram compostos por tr?s operadores de recombina??o (C1, C2, C3), dois operadores de constru??o de solu??o, um operador de muta??o e um operador de busca local. Por fim, um extenso processo de experimenta??o dos algoritmos ? realizado. Este processo possui tr?s etapas: a primeira etapa consistiu de ajustar os par?metros que cada algoritmo necessita, neste caso o ajuste de par?metro foi realizado para todas as vers?es do SPEA2, NSGA2 e GRASP; A segunda etapa consistiu de verificar, para cada algoritmo, qual a melhor vers?o. Por fim, as melhores vers?es de cada algoritmo, no total 3 vers?es, foram comparadas entre si visando identificar qual o melhor algoritmo dentre todos. Os algoritmos foram avaliados com base nos indicadores de qualidade Hypervolume e Epsilon Multiplicativo. Os resultados dos experimentos foram avaliados atrav?s de testes estat?sticos n?o-param?tricos (teste de Mann-Whitney e teste de Friedman). A avalia??o dos resultados foi favor?ravel ao NSGA2-C2 segundo a metodologia de avalia??o utilizada

Submitted by Leonardo Cavalcante (leo.ocavalcante@gmail.com) on 2018-06-11T11:34:10Z No. of bitstreams: 1 Arquivototal.pdf: 1995596 bytes, checksum: bfcc1e1f3a77514dcbf7a8e4f5e4701b (MD5) / Made available in DSpace on 2018-06-11T11:34:10Z (GMT). No. of bitstreams: 1 Arquivototal.pdf: 1995596 bytes, checksum: bfcc1e1f3a77514dcbf7a8e4f5e4701b (MD5) Previous issue date: 2017-02-16 / Conselho Nacional de Pesquisa e Desenvolvimento Científico e Tecnológico - CNPq / The problem in investment portfolio selection consists in the allocation of resources to a finite number of assets, aiming, in its classic approach, to overcome a trade-off between the risk and expected return of the portfolio. This problem is one of the most important topics targeted at today’s financial and economic issues. Since the pioneering works of Markowitz, the issue is treated as an optimisation problem with the two aforementioned objectives. However, in recent years, various restrictions and additional risk measurements were identified in the literature, such as, for example, cardinality restrictions, minimum transaction lot and asset pre-selection. This practice aims to bring the issue closer to the reality encountered in financial markets. In that regard, this paper proposes a metaheuristic called Particle Swarm for the optimisation of several PSPs, in such a way that allows the resolution of the problem considering a set of restrictions chosen by the investor. / O problema de seleção de carteiras de investimentos (PSP) consiste na alocação de recursos a um número finito de ativos, objetivando, em sua abordagem clássica, superar um trade-off entre o retorno esperado e o risco da carteira. Tal problema ´e uma das temáticas mais importantes voltadas a questões financeiras e econômicas da atualidade. Desde os pioneiros trabalhos de Markowitz, o assunto é tratado como um problema de otimização com esses dois objetivos citados. Entretanto, nos últimos anos, diversas restrições e mensurações de riscos adicionais foram consideradas na literatura, como, por exemplo, restrições de cardinalidade, de lote mínimo de transação e de pré-seleção de ativos. Tal prática visa aproximar o problema da realidade encontrada nos mercados financeiros. Neste contexto, o presente trabalho propõe uma meta-heurística denominada Adaptive Non-dominated Sorting Multiobjective Particle Swarm Optimization para a otimização de vários problemas envolvendo PSP, de modo que permita a resolução do problema considerando um conjunto de restri¸c˜oes escolhidas pelo investidor.

L'amélioration de la gestion des ressources naturelles est nécessaire pour répondre aux nombreux enjeux liés à leur exploitation. Ce travail propose une méthodologie d'optimisation de leur gestion, appliquée au cas du cuivre en France. Quatre critères permettant de juger les stratégies de gestion ont été retenus : le coût, les impacts environnementaux, la consommation énergétique et les pertes de ressources. La première étape de cette méthodologie est l'analyse de la situation actuelle, grâce à une modélisation du cycle français du cuivre de 2000 à 2009. Cet examen a montré que la France importe la quasi-totalité de ses besoins sous forme de cuivre raffiné, et a une industrie de recyclage peu développée. Suite à ces premiers résultats, la problématique du traitement des déchets de cuivre, et notamment de leur recyclage, a été étudiée. Une stratégie de modélisation des flux recyclés, basée sur la construction de flowsheets, a été développée. La formulation mathématique générale du problème a ensuite été définie : il s'agit d'un problème mixte, non-linéaire et a priori multiobjectif, qui a une contrainte égalité forte (la conservation de la masse). Une étude des méthodes d'optimisation a conduit à choisir un algorithme génétique (AG). Une alternative a également été envisagée pour résoudre le problème multiobjectif par programmation linéaire en le linéarisant "sous contrainte". Ce travail a mis en évidence la nécessité de développer une filière de recyclage efficace des déchets électriques et électroniques en France. Il a de plus montré que le cuivre contenu dans les déchets ne permet pas de couvrir la demande et qu'il est nécessaire d'importer du cuivre, de préférence sous forme de débris. / Improving the natural resources management is necessary to address the many issues related to their exploitation. This work proposes an optimization methodology for their management, applied to the case of copper in France. Four criteria are identified to assess management strategies: cost, environmental impacts, energy consumption and resource losses. The first step of this methodology is the analysis of the current situation, by modelling the French copper cycle from 2000 to 2009. This analysis showed that France imports almost all of its needs as refined copper, and has an underdeveloped recycling industry. Following these initial results, the problematic of copper wastes, including recycling, has been investigated. A recycled flow modelling strategy has been developed, based on the construction of flowsheets. The general mathematical formulation of the problem is then defined. It is a non-linear, mixed and a priori multiobjective problem, with a strong equality constraint (mass conservation). A review of optimization methods has led to choose a genetic algorithm (GA). An alternative was also proposed to solve the multiobjective problem with linear programming, by linearizing it under constraint. This work has highlighted the necessity of developing an effective recycling field of wastes from electric and electronic equipment in France. It also showed that the copper contained in wastes does not meet the demand, so that France needs to import copper, preferably as scraps.

Nous étudions la distillation extractive continue de mélanges azéotropiques à temperature de bulle minimale avec un entraineur lourd (classe 1.0-1a) avec comme exemples les mélanges acétone-méthanol avec l’eau et DIPE-IPA avec le 2-méthoxyethanol. Le procédé inclut les colonnes de distillation extractive et de régénération de l’entraineur en boucle ouverte et en boucle fermée. Une première stratégie d’optimisation consiste à minimiser la fonction objectif OF en cherchant les valeurs optimales du débit d’entraineur FE, les positions des alimentations en entraineur et en mélange NFE, NFAB, NFReg, les taux de reflux R1, R2 et les débits de distillat de chaque colonne D1, D2. OF décrit la demande en énergie par quantité de distillat et tient compte des différences de prix entre les utilités chaudes et froides et entre les deux produits. La deuxième stratégie est une optimisation multiobjectif qui minimise OF, le coût total annualisé (TAC) et maximise deux nouveaux indicateurs thermodynamiques d’efficacité de séparation extractive totale Eext et par plateau eext. Ils décrivent la capacité de la section extractive à séparer le produit entre le haut et le bas de la section extractive. L’analyse thermodynamique des réseaux de courbes de résidu ternaires RCM et des courbes d’isovolatilité montre l’intérêt de réduire la pression opératoire dans la colonne extractive pour les séparations de mélanges 1.0-1a. Une pression réduite diminue la quantité minimale d’entraineur et accroît la volatilité relative du mélange binaire azéotropique dans la région d’opération de la colonne extractive. Cela permet d’utiliser un taux de reflux plus faible et diminue la demande énergétique. La première stratégie d’optimisation est conduite avec des contraintes sur la pureté des produits avec les algorithmes SQP dans les simulateurs Aspen Plus ou Prosim Plus en boucle ouverte. Les variables continues optimisées sont : R1, R2 et FE (étape 1). Une étude de sensibilité permet de trouver les valeurs de D1, D2 (étape 2) et NFE, NFAB, NFReg (étape 3), tandis l’étape 1 est faite pour chaque jeu de variables discrètes. Enfin le procédé est resimulé en boucle fermée et TAC, Eext et eext sont calculés (étape 4). Les bilans matières expliquent l’interdépendance des débits de distillats et des puretés des produits. Cette optimisation permet de concevoir des procédés avec des gains proches de 20% en énergie et en coût. Les nouveaux procédés montrent une amélioration des indicateurs Eext et eext. Afin d’évaluer l’influence de Eext et eext sur la solution optimale, la seconde optimisation multiobjectif est conduite. L’algorithme génétique est peu sensible à l’initialisation, permet d’optimiser les variables discrètes N1, N2 et utilise directement le shéma de procédé en boucle fermée. L’analyse du front de Pareto des solutions met en évidence l’effet de FE/F et R1 sur TAC et Eext. Il existe un Eext maximum (resp. R1 minimum) pour un R1 donné (resp. Eext). Il existe aussi un indicateur optimal Eext,opt pour le procédé optimal avec le plus faible TAC. Eext,opt ne peut pas être utilisé comme seule fonction objectif d’optimisation mais en complément des autres fonctions OF et TAC. L’analyse des réseaux de profils de composition extractive explique la frontière du front de Pareto et pourquoi Eext augmente lorsque FE diminue et R1 augmente, le tout en lien avec le nombre d’étage. Visant à réduire encore TAC et la demande énergétique nous étudions des procédés avec intégration énergétique double effet (TEHI) ou avec des pompes à chaleur (MHP). En TEHI, un nouveau schéma avec une intégration énergétique partielle PHI réduit le plus la demande énergétique. En MHP, la recompression partielle des vapeurs VRC et bottom flash partiel BF améliorent les performances de 60% et 40% respectivement. Au final, le procédé PHI est le moins coûteux tandis que la recompression totale des vapeurs est la moins énergivore. / We study the continuous extractive distillation of minimum boiling azeotropic mixtures with a heavy entrainer (class 1.0-1a) for the acetone-methanol with water and DIPE-IPA with 2-methoxyethanol systems. The process includes both the extractive and the regeneration columns in open loop flowsheet and closed loop flowsheet where the solvent is recycled to the first column. The first optimization strategy minimizes OF and seeks suitable values of the entrainer flowrate FE, entrainer and azeotrope feed locations NFE, NFAB, NFReg, reflux ratios R1, R2 and both distillates D1, D2. OF describes the energy demand at the reboiler and condenser in both columns per product flow rate. It accounts for the price differences in heating and cooling energy and in product sales. The second strategy relies upon the use of a multi-objective genetic algorithm that minimizes OF, total annualized cost (TAC) and maximizes two novel extractive thermodynamic efficiency indicators: total Eext and per tray eext. They describe the ability of the extractive section to discriminate the product between the top and to bottom of the extractive section. Thermodynamic insight from the analysis of the ternary RCM and isovolatility curves shows the benefit of lowering the operating pressure of the extractive column for 1.0-1a class separations. A lower pressure reduces the minimal amount of entrainer and increases the relative volatility of original azeotropic mixture for the composition in the distillation region where the extractive column operates, leading to the decrease of the minimal reflux ratio and energy consumption. The first optimization strategy is conducted in four steps under distillation purity specifications: Aspen Plus or Prosim Plus simulator built-in SQP method is used for the optimization of the continuous variables: R1, R2 and FE by minimizing OF in open loop flowsheet (step 1). Then, a sensitivity analysis is performed to find optimal values of D1, D2 (step 2) and NFE, NFAB, NFReg (step 3), while step 1 is done for each set of discrete variables. Finally the design is simulated in closed loop flowsheet, and we calculate TAC and Eext and eext (step 4). We also derive from mass balance the non-linear relationships between the two distillates and how they relate product purities and recoveries. The results show that double digit savings can be achieved over designs published in the literature thanks to the improving of Eext and eext. Then, we study the influence of the Eext and eext on the optimal solution, and we run the second multiobjective optimization strategy. The genetic algorithm is usually not sensitive to initialization. It allows finding optimal total tray numbers N1, N2 values and is directly used with the closed loop flow sheet. Within Pareto front, the effects of main variables FE/F and R1 on TAC and Eext are shown. There is a maximum Eext (resp. minimum R1) for a given R1 (resp. Eext). There exists an optimal efficiency indicator Eext,opt which corresponds to the optimal design with the lowest TAC. Eext,opt can be used as a complementary criterion for the evaluation of different designs. Through the analysis of extractive profile map, we explain why Eext increases following the decrease of FE and the increase of R1 and we relate them to the tray numbers. With the sake of further savings of TAC and increase of the environmental performance, double-effect heat integration (TEHI) and mechanical heat pump (MHP) techniques are studied. In TEHI, we propose a novel optimal partial HI process aiming at the most energy saving. In MHP, we propose the partial VRC and partial BF heat pump processes for which the coefficients of performance increase by 60% and 40%. Overall, optimal partial HI process is preferred from the economical view while full VRC is the choice from the environmental perspective.

Les avantages que présentent les systèmes électriques par rapport aux autres systèmes (mécaniques, hydrauliques et pneumatiques) ont permis d’intensifier l’électrification des systèmes embarqués à bord des aéronefs : c’est le concept d’avion plus électrique. Dans ce contexte, l’approche de conception intégrée par optimisation (CIO) de ces systèmes s’avère aujourd’hui une nécessité pour pouvoir répondre aux exigences en termes d’efficacité énergique, de fiabilité et de masse... Dans cette thèse, nous avons appliqué la CIO à une chaine de conversion électromécanique utilisée dans le système de conditionnement d’air d’un avion. Deux objectifs sont ciblés : la minimisation de la masse du système et l’augmentation de son efficacité énergétique. Ces objectifs sont intégrés à diverses contraintes hétérogènes, allant de la qualité réseau au respect de la mission de vol dans le plan couple – vitesse, en passant par la thermique,… Compte tenu de la complexité du système étudié et de son caractère multidisciplinaire, des approches de conception par optimisation dites « MDO » (pour Multidisciplinary Design Optimization) sont étudiées. En effet, au delà des compétences physiques et techniques, la conception intégrée par optimisation des systèmes complexes nécessite des efforts supplémentaires en termes de méthodologies de conception. Nous avons présenté dans cette thèse trois approches : Approches mono-niveau : séquentielle et globale ; Approche multi-niveaux, couplant niveaux système et niveau constituants (filtre, onduleur, machine) ; des formulations adaptées à notre problème de conception sont présentées afin de résoudre les problèmes liés aux optimisations mono-niveau. Les performances des différentes approches de conception sont présentées analysées et comparées. Les résultats obtenus montrent clairement les avantages que présente la formulation multi-niveaux par rapport aux approches classiques de conception. / The benefits of electrical systems compared to other systems (mechanical, hydraulic and pneumatic) are a serious motivation for the electrification of embedded systems in “more electric aircraft”. In this framework, the integrated optimal design of these systems appears necessary to meet requirements in terms of efficiency, reliability and weight reduction. In this thesis, we have applied the integrated optimal design to an electromechanical system used in the air conditioning system of a more electric aircraft. Two objectives are targeted: the minimization of the system weight and the increase of its efficiency. Both objectives are integrated with several heterogeneous constraints, from network quality till flight mission fulfilment in the torque vs speed plan. Because of the complexity of the studied system and its multidisciplinary nature, "MDO" approaches (for multidisciplinary Design Optimization) are studied. In fact, beyond physical and technical skills, integrated optimal design of complex systems requires additional efforts in terms of design methodologies. Three approaches are presented in this thesis: One-level Approaches: sequential and global; Multilevel approach, coupling “system” level with “device” level (filter, inverter, electric machine); a set of formulations adapted to our design problem are presented to solve the issues associated to the one-level approaches. The performance of these design approaches are presented, analyzed and compared. The results clearly show the advantages that involves multilevel formulation compared to conventional design approaches.